Competition, Benchmarking and Incentives

Stream Director: Carol Propper

Current Research

Targets and Waiting Times: Exploiting a Quasi-Experiment to evaluate the use of targets in the provision of health care in the UK - (Complete)

Funder: ESRC
Staff: Carol Propper, Frank Windmeijer, Matthew Sutton (University of Aberdeen

The aim is to exploit policy variation in the use of waiting times targets in Scotland and England to evaluate the usefulness of targets as a means of increasing NHS productivity.

Pay, staffing and NHS output

Funder: ESRC
Staff: Carol Propper, John Van Reenen (LSE)

This research examines the impact of staffing and wage differentials on the output of NHS hospitals. We have found that both staffing matters (the proportion of qualified staff employed in a hospital raises quality and quantity of output) and that outside wages matter: hospitals operating in markets where outside wages for nurses are higher have lower quality (higher death rates) and lower volume of output. Work this year has focused on the route by which lower wages translate into poorer outputs occur. We have identified the use of agency nurses as a possible source: we are currently subjecting these findings to a battery of robustness checks.

Do Waiting time targets work?

Funder: ESRC
Staff: Carol Propper, Matt Sutton (Aberdeen)

In 2001 the English NHS introduced a policy of targets designed to reduce waiting times. We have examined whether this policy worked by comparing waiting times in England with those in Scotland, where no policy was implemented. We find that the policy did work. Waiting times in England for elective hospital care fell after implemented of the policy relative to Scotland. They fell not only for those waits which were the focus of policy but also for shorter ones too. Thus the policy appears to have achieved its desired end.

Travel and choice in the NHS

Funder: Leverhulme / ESRC
Staff: Carol Propper, George Leckie, Michael Damiani (Kings Fund) and Jennifer Dixon (Kings Fund)

Current policy in the NHS in England seeks to extend the choice of provider of care for patients. Two years ago we began a project examining travel patterns prior to the introduction of the current choice based policies. We found that there is a strong gradient in travel by ward deprivation: while on average individuals living in poorer areas travel only a little less than all others, very few people in these areas travel long distances. We also find that these differences persist within major specialites. We are now investigating what the sources of these differences are focusing on the role played by GPs and hospitals.

Exit, voice and public service delivery

Funder: ESRC

Consumer choice – or exit – is currently a key element of public service reform, and the current policy discourse suggests that choice and voice will complement each other to create user-driven pressures for improvements in public service quality. This research scrutinises the accuracy of this discourse by going back to Hirschman’s thesis and investigating the conditions under which exit and voice are complementary and those under which they are not. The policy implications are then considered within the institutional framework of the English education system.

Contextual value added, league tables and the basis for school choice

Funder: ESRC

Parental choice among schools in England is informed by annually published school performance (league) tables. In this project we use UK government administrative data to replicate CVA in order to investigate the extent to which the current league tables provide the information necessary to support parental choice on the basis of school effectiveness. We find that while CVA does provide a more accurate measure of school performance, school rankings based on CVA are largely meaningless: almost half of English secondary schools are indistinguishable from the national average.

School quality and neighbourhood formation

Funder: ESRC

In on-going work, we investigate the relationship between school quality and families’ decisions to move house. We show that families are more likely to move away from schools with lower published results, and also show a differential response between poor and non-poor families. We also show that the strength of the relationship depends on factors in the local education markets.

Incentives in teams

Funder: ESRC

We have undertaken over the years a series of projects on the impact of team incentives on output in the public sector. This year we have revisited research on a scheme implemented in Custom and Excise to examine whether the scheme increased output by encouraging extra effort or by strategic reallocation of more efficient workers to tasks which were rewarded under the scheme. We find evidence of considerable reallocation of workers that was associated with better pay-off for the team members.

Incentive mechanisms to promote cooperation

Funder: ESRC

We consider two mechanisms proposed by the theoretical literature to solve the free-rider problem in public goods situations, which, from a theoretical point of view, produce the same outcome. We test this prediction using laboratory experiments. We do find that both mechanisms increase contributions to the public good, but that in some cases, the outcomes are significantly different.

Choice in Public services

Funder: ESRC
Staff: Simon Burgess, Carol Propper, Deborah Wilson

Building on last year’s detailed review of the economics literature, we have provided an analysis of (i) the current NHS choice based policy and the implications for regulation of the evolving market in healthcare, and (ii) the strengthening of school choice in the Education bill.

School choice in England

Funder: ESRC/Leverhulme
Staff: Simon Burgess, Adam Briggs, Deborah Wilson

We use PLASC/NPD with detailed data on pupils’ test scores, locations, pupil characteristics and school characteristics. We provide evidence on school commutes, on distances to good schools, and on who goes to their local school. The paper shows significant differences in school commute times, and availability of school choice by geography, by ethnicity and by deprivation.

School assignment, school choice and social mobility

We study in greater depth the mechanisms underlying the sorting of children across different schools within an area. We find that poor children are less likely to go to good schools. Some of this difference between poor and rich children is accounted for by location, but even controlling for exact residential location, the characteristics of the child and of her neighbourhood, poor children are less likely to go to good schools.

The impact of school competition on test scores

It has proved difficult to isolate exogenous differences in the degree of competition faced by schools. We exploit a local government reorganisation to provide identification. We measure the impact of this fall in competition on pupil progress, controlling for initial pupil attainment, other pupil characteristics and school effects. Our results show no significant effect of the decline in competition on pupil progress.

Student attainment and progress, and the role of geography

We investigated a series of multilevel models to analyze the role of geography in pupil attainment and progress at Key Stage 4. For a sample of four Local Unitary Authorities, we examined models allowing for Secondary and Primary school effects and for geography level effects: effects of Output Areas and Lower and Middle Layers of Super Output Areas. We found that progress is less dependent on geography than the level of attainment.

Incentive Design and public service outcomes - ongoing

Funder: ESRC / Leverhulme
Staff: Simon Burgess, Carol Propper, Marisa Ratto, Stephanie von Hinke Kessler Scholder, Elisa Sicuri

We are continuing to exploit our data collection around the introduction of performance pilots in some large government agencies. We are completing work on the ways that incentives work through affecting both workers and managers. The results for the agency we study show that the appropriate allocation of staff to tasks by managers is more important than greater effort by workers. New research on the theory of performance monitoring includes analysing the implications of performance measures over different tasks, and the design of incentives when workers work in teams.

Executive pay and performance - complete

Funder: Leverhulme
Staff: Paul Gregg, Ian Tonks (Exeter)

This paper examines the relationship between executive cash compensation and company performance for a sample of large UK companies over the period 1994-2002. This relationship is examined against a background of a series of reports into corporate governance mechanisms in UK companies. We show that base pay compensation of UK executives has increased substantially over this period, and we provide evidence on the movement in the pay-performance sensitivity over time. We identify an asymmetric relationship between pay and performance: in years and for companies in which stock returns are relatively high, pay-performance elasticities are high, but we find that executive pay is less sensitive to performance in those cases when stock returns are low. This suggests that overall there is little relationship between pay and performance. We also explore the heterogeneity of the pay-performance relationship across firms, and find that board structure, firm size, industry and firm risk are all significant determinants of executive compensation.

Incentive Design, Cream Skimming and public service motivation - On-going

Funder: ESRC
Staff: Simon Burgess, Carol Propper, Marisa Ratto, Stephanie von Hinke Kessler Scholder

We have analysed the implied prioritisation of cases by Jobcentre staff when part of an incentive scheme. This relates to the debate on "cream skimming" - whether staff prioritise easier cases to deal with. With our data from Jobcentre Plus, we know the rates at which different categories of job seeker leave Jobcentre Plus's register. These categories include long-term and short-term unemployed, disabled persons and lone mothers. We compare exit rates for these groups with and without the use of incentives, and across offices in the incentive scheme and out. This research is nearing completion.

The public-private pay gap - complete

Funder: Leverhulme
Staff: Fabien Postel-Vinay, Helene Turon

The existing literature on inequality between private and public sectors focuses on cross-section differences in earnings levels. A more general way of looking at inequality between sectors is to recognize that forward-looking agents will care about income and job mobility too. We show that these are substantially different between the two sectors. Using data from the BHPS, we estimate a model of income and employment dynamics over seven years. We allow for unobserved heterogeneity in the propensity to be unemployed or employed in either job sector and in terms of the income process. We then combine the results into lifetime values of jobs in either sector and carry out a cross-section comparative analysis of these values. We have four main findings. First focusing on cross-sector differences in terms of the income process only, we detect a positive average public premium both in income flows and in the present discounted sum of future income flows. Second, we argue that income inequality is lower but more persistent in the public sector, as most of the observed relative cross-sectional income compression in the public sector is due to a lower variance of the transitory component of income. Third, when taking job mobility into account, the lifetime public premium is essentially zero for workers that we categorize as "high-employability" individuals, suggesting that the UK labor market is sufficiently mobile to ensure a rapid allocation of workers into their "natural" sector. Fourth, we find some evidence of job queuing for public sector jobs among "low-employability" workers.

Choice and Sorting in English secondary schools

Funder: Leverhulme
Staff: Simon Burgess, Carol Propper, Deborah Wilson, Brendon McConnell

We establish that post-residential school choice is an important component of the overall schooling decision. We show that there is a difference in the school-neighbourhood sorting relationship between areas that operate under different student-to-school assignment rules.

The impact of school league tables on pupil attainment - complete

Funder: Leverhulme
Staff: Simon Burgess, Carol Propper, Helen Slater, Deborah Wilson

In 1988 the UK government introduced greater accountability into the English state school sector. But the information that schools are required to make public on their pupil achievement is only partial. We examine whether accountability measures based on a partial summary of student achievement influence the distribution of student achievement. Since school ratings only incorporate test results via pass rates, schools have incentives to improve the performance of students who are on the margin of meeting these standards, to the detriment of very low achieving or high achieving pupils. Using pupil level data for a cohort of all students in secondary public sector schools in England, we find that this policy reduces the educational gains and exam performance in high stakes exams of very low ability students.